Flamingo Origins

[Peter Nyikos]

Plastic flamingos are a popular lawn decoration where I live, and when
I saw three of them clustered around a mailbox, I got to wondering
where the live birds can be found in the wild.

Much to my surprise, my only map that showed their USA distribution
had them in the extreme south of Florida -- the Florida keys and not much
else. And it suggested that they might all be escapees.

_Peterson's Field Guides_ had this to say about the range of the one species
involved, the American flamingo *Phoenicopterus ruber*:

Range: W. Indies, Yucatan, Galapagos Is. East: Occasional on the
Florida coast; accidental elsewhere. It is difficult to determine
whether vagrants are truly wild or escapees. Hialeah Race Track,
Bok Sanctuary, Busch Gardens, etc., may be the sources of strays.

So I got to wondering where the flamingo might have originated.
The oldest fossils, according to the Wikipedia, are from the
Old World:

Prehistoric species of flamingo:[citation needed]
Phoenicopterus croizeti (Middle Oligocene – Middle Miocene of C Europe)
...
Phoenicopterus eyrensis (Late Oligocene of South Australia)
Phoenicopterus novaehollandiae (Late Oligocene of South Australia)
https://en.wikipedia.org/wiki/Flamingo

The other "prehistoric" species are from the Pliocene or later. None are
listed as being from South America, Asia or Africa, the places where
flamingos are most widespread today. And none of the extant ones are
to be found wild in Australia at all!

Of course, "prehistoric" could be used by the author of the entry
to mean "known only from fossils" and perhaps some of the extant
species have a fossil record. I hope John Harshman, our resident expert
on ornithology, can clarify this issue for us.

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer --
University of South Carolina

[Peter Nyikos]

On Monday, August 22, 2016 at 8:56:34 PM UTC-4, Peter Nyikos wrote:

Besides geographical origins of flamingos, it is also very much on-topic
to talk about their phylogenetic origins. There have been a number of
hypotheses down through the years:

Traditionally, the long-legged Ciconiiformes, probably a paraphyletic
assemblage, have been considered the flamingos' closest relatives
and the family was included in the order. Usually the ibises and
spoonbills of the Threskiornithidae were considered their closest
relatives within this order.

Earlier genetic studies, such as those of Charles Sibley and colleagues,
also supported this relationship.[3] Relationships to the waterfowl were
considered as well,[4] especially as flamingos are parasitized by feather
lice of the genus Anaticola, which are otherwise exclusively found on
ducks and geese.[5] The peculiar presbyornithids were used to argue
for a close relationship between flamingos, waterfowl, and waders.[6]
A 2002 paper concluded they are waterfowl,[7] but a 2014 comprehensive
study of bird orders found that flamingos and grebes are not waterfowl,
but rather are part of Columbea along with doves, sandgrouse,
and mesites.[8]

https://en.wikipedia.org/wiki/Flamingo

[8] Jarvis, E.D.; et al. (2014). "Whole-genome analyses resolve early
branches in the tree of life of modern birds". Science. 346 (6215): 1320–
1331. doi:10.1126/science.1253451. PMC 4405904free to read. PMID 25504713.

As is so often the case, the latest analysis is made to sound like it
settles the question of phylogeny, but I have to wonder. "et al"
includes something like seventy (!) co-authors, which is not always
conducive to a consistent article. That aside, the abstract also
sounds a sort of cautionary note:

Even with whole genomes, some of the earliest branches in Neoaves proved
challenging to resolve, which was best explained by massive protein-coding
sequence convergence and high levels of incomplete lineage sorting that
occurred during a rapid radiation after the Cretaceous-Paleogene mass
extinction event about 66 million years ago.

Then too, I could see no sign of fossil evidence being used in this
study, although a couple of references do seem to center on the fossil
record.

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer --
University of South Carolina

http://people.math.sc.edu/nyikos/

[rsNorman]

Peter Nyikos <nyi...@bellsouth.net> Wrote in message:

> Plastic flamingos are a popular lawn decoration where I live, and when
> I saw three of them clustered around a mailbox, I got to wondering
> where the live birds can be found in the wild.
>
> Much to my surprise, my only map that showed their USA distribution
> had them in the extreme south of Florida -- the Florida keys and not much
> else. And it suggested that they might all be escapees.
>
> _Peterson's Field Guides_ had this to say about the range of the one species
> involved, the American flamingo *Phoenicopterus ruber*:
>
> Range: W. Indies, Yucatan, Galapagos Is. East: Occasional on the
> Florida coast; accidental elsewhere. It is difficult to determine
> whether vagrants are truly wild or escapees. Hialeah Race Track,
> Bok Sanctuary, Busch Gardens, etc., may be the sources of strays.
>
> So I got to wondering where the flamingo might have originated.
> The oldest fossils, according to the Wikipedia, are from the
> Old World:
>
> Prehistoric species of flamingo:[citation needed]

> Phoenicopterus croizeti (Middle Oligocene ? Middle Miocene of C Europe)

> ...
> Phoenicopterus eyrensis (Late Oligocene of South Australia)
> Phoenicopterus novaehollandiae (Late Oligocene of South Australia)
> https://en.wikipedia.org/wiki/Flamingo
>
> The other "prehistoric" species are from the Pliocene or later. None are
> listed as being from South America, Asia or Africa, the places where
> flamingos are most widespread today. And none of the extant ones are
> to be found wild in Australia at all!
>
> Of course, "prehistoric" could be used by the author of the entry
> to mean "known only from fossils" and perhaps some of the extant
> species have a fossil record. I hope John Harshman, our resident expert
> on ornithology, can clarify this issue for us.
>

Did you not see Feduccia's paper on evolution of flamingos? I
can't give you the web site using my current hardware but it
is
Relationships and Evolution of Flamingos (Aves: Phoenicopteridae)
Storrs L. Olson and Alan Feduccia
Smithsonian Contributions to Zoology 316 1980

It should show up in a Google search for "evolution of flamingos"



--


----Android NewsGroup Reader----
http://usenet.sinaapp.com/

[John Harshman]

I don't know that much about Pleistocene fossils. But genetically,
extant species of flamingos are all very similar. I would estimate a
Pleistocene divergence. More interesting to me is their divergence from
their living sister group, grebes. Gerald Mayr had a paper about this,
including a possibly basal flamingo:

http://www.senckenberg.de/files/content/forschung/abteilung/terrzool/ornithologie/flamingo.pdf

[Peter Nyikos]

Thanks, I found it in Google Scholar. The whole book of 73 pages is
available in PDF form. It'll take a while to get the feel of it.

The book is listed in the extensive biography of the *Science* article
that I quoted from in my second post to this thread. Reference [3] there
is to a 1995 paper by Feduccia:

A. Feduccia, "Explosive evolution in tertiary birds and mammals."
Science 267, 637– 638 (1995). doi:10.1126/science.267.5198.637
pmid:17745839 FREE Full TextGoogle Scholar

Peter Nyikos
Professor of Mathematics -- standard disclaimer --
University of South Carolina

[John Harshman]

On 8/22/16 6:35 PM, Peter Nyikos wrote:
> On Monday, August 22, 2016 at 8:56:34 PM UTC-4, Peter Nyikos wrote:
>
> Besides geographical origins of flamingos, it is also very much on-topic
> to talk about their phylogenetic origins. There have been a number of
> hypotheses down through the years:
>
> Traditionally, the long-legged Ciconiiformes, probably a paraphyletic
> assemblage,

Polyphyletic, ackshully.

> have been considered the flamingos' closest relatives

> but a 2014 comprehensive
> study of bird orders found that flamingos and grebes are not waterfowl,
> but rather are part of Columbea along with doves, sandgrouse,
> and mesites.[8]

The flamingos & grebes thing is true. The "not waterfowl" thing is
certainly true. Relationship to Columbea has been replicated in a couple
of other studies and might also be true.

> https://en.wikipedia.org/wiki/Flamingo
>
> [8] Jarvis, E.D.; et al. (2014). "Whole-genome analyses resolve early
> branches in the tree of life of modern birds". Science. 346 (6215): 1320–
> 1331. doi:10.1126/science.1253451. PMC 4405904free to read. PMID 25504713.
>
> As is so often the case, the latest analysis is made to sound like it
> settles the question of phylogeny, but I have to wonder. "et al"
> includes something like seventy (!) co-authors, which is not always
> conducive to a consistent article.

Turns out much of it can be supported by independent analyses.

> That aside, the abstract also
> sounds a sort of cautionary note:
>
> Even with whole genomes, some of the earliest branches in Neoaves proved
> challenging to resolve, which was best explained by massive protein-coding
> sequence convergence and high levels of incomplete lineage sorting that
> occurred during a rapid radiation after the Cretaceous-Paleogene mass
> extinction event about 66 million years ago.
>
> Then too, I could see no sign of fossil evidence being used in this
> study, although a couple of references do seem to center on the fossil
> record.

I see no problem with not using fossil evidence in a genomic study.

[John Harshman]

Was that an attempt at tweaking Peter's nose? Or did you not read that
publication? It's notorious for trying to derive flamingos from one
particular species of extant recurvirostrid.

[John Harshman]

"Transitional shorebirds". Snort.

[rsNorman]

John Harshman <jhar...@pacbell.net> Wrote in message:

It was gentle tweaking only to chide him on not trying a Google
search. That the monograph was by Feduccia was a bonus.


I only skimmed that 80+ page thing and did notice the frequent
occurrence of the Australian Banded Stilt. I have no reason to
doubt that flamingos might belong to the Charadriiformes
although
you might.

[John Harshman]

It was a ridiculous suggestion at the time, and it's much more
ridiculous in light of current data. Here's the real paper on flamingo
relationships:

https://www.researchgate.net/publication/11911863_Convergence_and_divergence_in_the_evolution_of_aquatic_birds

And here's a web site that discusses some of the evidence:

http://www.10000birds.com/exploring-the-relationship-between-flamingos-and-grebes-the-wonderful-birds.htm

[rsNorman]

This is your area. You know, I don't.

My excuse is "I saw it on the internet."

[Peter Nyikos]

On Monday, August 22, 2016 at 10:17:30 PM UTC-4, John Harshman wrote:
> On 8/22/16 6:35 PM, Peter Nyikos wrote:

> > Besides geographical origins of flamingos, it is also very much on-topic
> > to talk about their phylogenetic origins. There have been a number of
> > hypotheses down through the years:
> >
> > Traditionally, the long-legged Ciconiiformes, probably a paraphyletic
> > assemblage,
>
> Polyphyletic, ackshully.

What does it include? Jarvis's article, referenced below, makes
no mention of them.

> > have been considered the flamingos' closest relatives
>
> > but a 2014 comprehensive
> > study of bird orders found that flamingos and grebes are not waterfowl,
> > but rather are part of Columbea along with doves, sandgrouse,
> > and mesites.[8]
>
> The flamingos & grebes thing is true. The "not waterfowl" thing is
> certainly true.

I would have thought "waterfowl" included grebes (also loons, etc.)
but according to the Merriam-Webster dictionary, it only includes
water birds that are hunted for sport when used in the plural, as here.

Looks like hunting grebes never really caught on.

> Relationship to Columbea has been replicated in a couple
> of other studies and might also be true.

"might"? why so much less sure than with grebes?

> > https://en.wikipedia.org/wiki/Flamingo
> >
> > [8] Jarvis, E.D.; et al. (2014). "Whole-genome analyses resolve early
> > branches in the tree of life of modern birds". Science. 346 (6215): 1320–
> > 1331. doi:10.1126/science.1253451. PMC 4405904free to read. PMID 25504713.

This url takes you right there:
http://science.sciencemag.org/content/346/6215/1320.full

> > As is so often the case, the latest analysis is made to sound like it
> > settles the question of phylogeny, but I have to wonder. "et al"
> > includes something like seventy (!) co-authors, which is not always
> > conducive to a consistent article.

The first sentence in the abstract is already imprecise:

To better determine the history of modern birds, we performed a
genome-scale phylogenetic analysis of 48 species representing
all orders of Neoaves using phylogenomic methods created to handle
genome-scale data.

It also includes some representatives of some (not all) orders of
Palaeognathae and Galloanserae, presumably to root the tree.
You can find them in Fig. 1:
https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg

> Turns out much of it can be supported by independent analyses.
>
> > That aside, the abstract also
> > sounds a sort of cautionary note:
> >
> > Even with whole genomes, some of the earliest branches in Neoaves proved
> > challenging to resolve, which was best explained by massive protein-coding
> > sequence convergence and high levels of incomplete lineage sorting that
> > occurred during a rapid radiation after the Cretaceous-Paleogene mass
> > extinction event about 66 million years ago.
> >
> > Then too, I could see no sign of fossil evidence being used in this
> > study, although a couple of references do seem to center on the fossil
> > record.
>
> I see no problem with not using fossil evidence in a genomic study.

Given the "massive protein-coding sequence convergence," it seems
prudent to make sure the fossil evidence is not at odds with the
molecular.

Peter Nyikos
Professor, Department of Mathematics
University of South Carolina -- standard disclaimer --

[Peter Nyikos]

On Monday, August 22, 2016 at 10:19:53 PM UTC-4, John Harshman wrote:
> On 8/22/16 6:42 PM, rsNorman wrote:
> > Peter Nyikos <nyi...@bellsouth.net> Wrote in message:

Perhaps Richard was under the mistaken impression that I am partial
to Feduccia. I can't understand why else he would pick such an old (1980)
treatise. The Jarvis article alone has 277 references, most of them
much more recent.

[John Harshman]

On 8/23/16 6:15 PM, Peter Nyikos wrote:
> On Monday, August 22, 2016 at 10:17:30 PM UTC-4, John Harshman wrote:
>> On 8/22/16 6:35 PM, Peter Nyikos wrote:
>
>>> Besides geographical origins of flamingos, it is also very much on-topic
>>> to talk about their phylogenetic origins. There have been a number of
>>> hypotheses down through the years:
>>>
>>> Traditionally, the long-legged Ciconiiformes, probably a paraphyletic
>>> assemblage,
>>
>> Polyphyletic, ackshully.
>
> What does it include? Jarvis's article, referenced below, makes
> no mention of them.

Traditionally, Ciconiiformes includes flamingos, herons, ibises, and
storks. In fact, it includes only the storks. Jarvis et al. didn't put a
stork in their analysis, so you won't see it there. But try Hackett et
al. 2008.

http://earlybird.biology.ufl.edu/publications/

>>> have been considered the flamingos' closest relatives
>>
>>> but a 2014 comprehensive
>>> study of bird orders found that flamingos and grebes are not waterfowl,
>>> but rather are part of Columbea along with doves, sandgrouse,
>>> and mesites.[8]
>>
>> The flamingos & grebes thing is true. The "not waterfowl" thing is
>> certainly true.
>
> I would have thought "waterfowl" included grebes (also loons, etc.)
> but according to the Merriam-Webster dictionary, it only includes
> water birds that are hunted for sport when used in the plural, as here.

As commonly used, it's a synonym for "anseriform" (minus screamers).

> Looks like hunting grebes never really caught on.
>
>> Relationship to Columbea has been replicated in a couple
>> of other studies and might also be true.
>
> "might"? why so much less sure than with grebes?

Flamangos and grebes is easy to get with not that much data, and most
individual genes will tell you that. It just happens that nobody had
ever looked until van Tuinen. The relationship to Columbea requires lots
and lots of sequence and careful analysis. More replication would be good.

>>> https://en.wikipedia.org/wiki/Flamingo
>>>
>>> [8] Jarvis, E.D.; et al. (2014). "Whole-genome analyses resolve early
>>> branches in the tree of life of modern birds". Science. 346 (6215): 1320–
>>> 1331. doi:10.1126/science.1253451. PMC 4405904free to read. PMID 25504713.
>
> This url takes you right there:
> http://science.sciencemag.org/content/346/6215/1320.full
>
>>> As is so often the case, the latest analysis is made to sound like it
>>> settles the question of phylogeny, but I have to wonder. "et al"
>>> includes something like seventy (!) co-authors, which is not always
>>> conducive to a consistent article.
>
> The first sentence in the abstract is already imprecise:
>
> To better determine the history of modern birds, we performed a
> genome-scale phylogenetic analysis of 48 species representing
> all orders of Neoaves using phylogenomic methods created to handle
> genome-scale data.

> It also includes some representatives of some (not all) orders of
> Palaeognathae and Galloanserae, presumably to root the tree.
> You can find them in Fig. 1:
> https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg

Is this something more than a nitpick? Do you draw some lesson?

>> Turns out much of it can be supported by independent analyses.
>>
>>> That aside, the abstract also
>>> sounds a sort of cautionary note:
>>>
>>> Even with whole genomes, some of the earliest branches in Neoaves proved
>>> challenging to resolve, which was best explained by massive protein-coding
>>> sequence convergence and high levels of incomplete lineage sorting that
>>> occurred during a rapid radiation after the Cretaceous-Paleogene mass
>>> extinction event about 66 million years ago.
>>>
>>> Then too, I could see no sign of fossil evidence being used in this
>>> study, although a couple of references do seem to center on the fossil
>>> record.
>>
>> I see no problem with not using fossil evidence in a genomic study.
>
> Given the "massive protein-coding sequence convergence," it seems
> prudent to make sure the fossil evidence is not at odds with the
> molecular.

So it seems to you. But in fact the best check would be different
molecular data, which in fact Jarvis et al. did. Would you really expect
every study to use all different sorts of data, gained in completely
different ways and demanding different sorts of expertise? Would you
also suggest that every paleo paper should feature a DNA sequence analysis?

[John Harshman]

If you were ever partial to Feduccia, reading that monograph ought to
cure you.

[Peter Nyikos]

On Monday, August 22, 2016 at 10:06:05 PM UTC-4, John Harshman wrote:
> On 8/22/16 5:56 PM, Peter Nyikos wrote:

> > So I got to wondering where the flamingo might have originated.
> > The oldest fossils, according to the Wikipedia, are from the
> > Old World:
> >
> > Prehistoric species of flamingo:[citation needed]
> > Phoenicopterus croizeti (Middle Oligocene – Middle Miocene of C Europe)
> > ...
> > Phoenicopterus eyrensis (Late Oligocene of South Australia)
> > Phoenicopterus novaehollandiae (Late Oligocene of South Australia)
> > https://en.wikipedia.org/wiki/Flamingo
> >
> > The other "prehistoric" species are from the Pliocene or later. None are
> > listed as being from South America, Asia or Africa, the places where
> > flamingos are most widespread today. And none of the extant ones are
> > to be found wild in Australia at all!
> >
> > Of course, "prehistoric" could be used by the author of the entry
> > to mean "known only from fossils" and perhaps some of the extant
> > species have a fossil record. I hope John Harshman, our resident expert
> > on ornithology, can clarify this issue for us.
>
> I don't know that much about Pleistocene fossils.

Why Pleistocene? Is that the earliest epoch from which fossils of
extant flamingo species are found?

> But genetically,
> extant species of flamingos are all very similar. I would estimate a
> Pleistocene divergence. More interesting to me is their divergence from
> their living sister group, grebes.

According to Wikipedia, the extant grebe genus *Podiceps* goes back
at least to the early Miocene, as far back as any true grebe goes.
But FWIW, the Jarvis article puts the divergence much further back,
to 55 mya.

Keep in mind that "molecular clock" estimates are much less
precise than phylogenetic estimates. It was only a few years ago
that the molecular estimates for the LCA of Placentalia were brought
into somewhat closer line with the fossil evidence.

> Gerald Mayr had a paper about this,
> including a possibly basal flamingo:
>
> http://www.senckenberg.de/files/content/forschung/abteilung/terrzool/ornithologie/flamingo.pdf

I note that he puts grebes and flamingos in a clade with loons
and penguins and away from every other bird analyzed, using
morphological evidence. That's a far cry from where Jarvis et. al.
put loons and penguins using molecular evidence. Compare Mayr's
"strict consensus tree" in his Figure 1 with Jarvis's Fig. 1:
https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer --
University of South Carolina

http://www.math.sc.edu/~nyikos

[John Harshman]

On 8/23/16 7:26 PM, Peter Nyikos wrote:
> On Monday, August 22, 2016 at 10:06:05 PM UTC-4, John Harshman wrote:
>> On 8/22/16 5:56 PM, Peter Nyikos wrote:
>
>>> So I got to wondering where the flamingo might have originated.
>>> The oldest fossils, according to the Wikipedia, are from the
>>> Old World:
>>>
>>> Prehistoric species of flamingo:[citation needed]
>>> Phoenicopterus croizeti (Middle Oligocene – Middle Miocene of C Europe)
>>> ...
>>> Phoenicopterus eyrensis (Late Oligocene of South Australia)
>>> Phoenicopterus novaehollandiae (Late Oligocene of South Australia)
>>> https://en.wikipedia.org/wiki/Flamingo
>>>
>>> The other "prehistoric" species are from the Pliocene or later. None are
>>> listed as being from South America, Asia or Africa, the places where
>>> flamingos are most widespread today. And none of the extant ones are
>>> to be found wild in Australia at all!
>>>
>>> Of course, "prehistoric" could be used by the author of the entry
>>> to mean "known only from fossils" and perhaps some of the extant
>>> species have a fossil record. I hope John Harshman, our resident expert
>>> on ornithology, can clarify this issue for us.
>>
>> I don't know that much about Pleistocene fossils.
>
> Why Pleistocene? Is that the earliest epoch from which fossils of
> extant flamingo species are found?

Dunno.

>> But genetically,
>> extant species of flamingos are all very similar. I would estimate a
>> Pleistocene divergence. More interesting to me is their divergence from
>> their living sister group, grebes.
>
> According to Wikipedia, the extant grebe genus *Podiceps* goes back
> at least to the early Miocene, as far back as any true grebe goes.
> But FWIW, the Jarvis article puts the divergence much further back,
> to 55 mya.

I'm talking about the divergence within crown flamingos.

> Keep in mind that "molecular clock" estimates are much less
> precise than phylogenetic estimates. It was only a few years ago
> that the molecular estimates for the LCA of Placentalia were brought
> into somewhat closer line with the fossil evidence.
>
>> Gerald Mayr had a paper about this,
>> including a possibly basal flamingo:
>>
>> http://www.senckenberg.de/files/content/forschung/abteilung/terrzool/ornithologie/flamingo.pdf
>
> I note that he puts grebes and flamingos in a clade with loons
> and penguins and away from every other bird analyzed, using
> morphological evidence. That's a far cry from where Jarvis et. al.
> put loons and penguins using molecular evidence. Compare Mayr's
> "strict consensus tree" in his Figure 1 with Jarvis's Fig. 1:
> https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg

Yes. Mayr's analysis is wrong about that. Jarvis is correct.

[Peter Nyikos]

So, why Pleistocene? Are there no "living fossils" among birds, that
"very similar" is enough to conclude such a late divergence?

> >> But genetically,
> >> extant species of flamingos are all very similar. I would estimate a
> >> Pleistocene divergence.

Over on a thread about squirrels that Oxyaena began, I quoted the following
from Carroll [1988]:

Miocene members of the genus *Sciurus* are so similar to
living squirrels that Emry and Thorington (1984) considered this
genus to be a "living fossil," despite its worldwide distribution
and great species diversity in the modern fauna.

Note, "diversity" not "disparity."

> >> More interesting to me is their divergence from
> >> their living sister group, grebes.
> >
> > According to Wikipedia, the extant grebe genus *Podiceps* goes back
> > at least to the early Miocene, as far back as any true grebe goes.
> > But FWIW, the Jarvis article puts the divergence much further back,
> > to 55 mya.
>
> I'm talking about the divergence within crown flamingos.

That was before you talked about "their divergence from
their living sister group, grebes." You want the crown
group including both flamingos and grebes for that.

> > Keep in mind that "molecular clock" estimates are much less
> > precise than phylogenetic estimates. It was only a few years ago
> > that the molecular estimates for the LCA of Placentalia were brought
> > into somewhat closer line with the fossil evidence.
> >
> >> Gerald Mayr had a paper about this,
> >> including a possibly basal flamingo:
> >>
> >> http://www.senckenberg.de/files/content/forschung/abteilung/terrzool/ornithologie/flamingo.pdf
> >
> > I note that he puts grebes and flamingos in a clade with loons
> > and penguins and away from every other bird analyzed, using
> > morphological evidence. That's a far cry from where Jarvis et. al.
> > put loons and penguins using molecular evidence. Compare Mayr's
> > "strict consensus tree" in his Figure 1 with Jarvis's Fig. 1:
> > https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg
>
> Yes. Mayr's analysis is wrong about that. Jarvis is correct.

Was Mayr inconsistent, or do you know of some independent analysis
(preferably morphological) that confirms Jarvis's correctness?

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer --

University of South Carolina, Columbia

[Peter Nyikos]

On Tuesday, August 23, 2016 at 9:43:47 PM UTC-4, John Harshman wrote:
> On 8/23/16 6:15 PM, Peter Nyikos wrote:
> > On Monday, August 22, 2016 at 10:17:30 PM UTC-4, John Harshman wrote:
> >> On 8/22/16 6:35 PM, Peter Nyikos wrote:
> >
> >>> Besides geographical origins of flamingos, it is also very much on-topic
> >>> to talk about their phylogenetic origins. There have been a number of
> >>> hypotheses down through the years:
> >>>
> >>> Traditionally, the long-legged Ciconiiformes, probably a paraphyletic
> >>> assemblage,
> >>
> >> Polyphyletic, ackshully.
> >
> > What does it include? Jarvis's article, referenced below, makes
> > no mention of them.
>
> Traditionally, Ciconiiformes includes flamingos, herons, ibises, and
> storks. In fact, it includes only the storks. Jarvis et al. didn't put a
> stork in their analysis, so you won't see it there. But try Hackett et
> al. 2008.
>
> http://earlybird.biology.ufl.edu/publications/

Thanks.

>
> >>> have been considered the flamingos' closest relatives
> >>
> >>> but a 2014 comprehensive
> >>> study of bird orders found that flamingos and grebes are not waterfowl,
> >>> but rather are part of Columbea along with doves, sandgrouse,
> >>> and mesites.[8]

<snip>

> >> Relationship to Columbea has been replicated in a couple
> >> of other studies and might also be true.
> >
> > "might"? why so much less sure than with grebes?
>
> Flamangos and grebes is easy to get with not that much data, and most
> individual genes will tell you that. It just happens that nobody had
> ever looked until van Tuinen. The relationship to Columbea requires lots
> and lots of sequence and careful analysis. More replication would be good.

Good summary. Thanks.

> >>> https://en.wikipedia.org/wiki/Flamingo
> >>>
> >>> [8] Jarvis, E.D.; et al. (2014). "Whole-genome analyses resolve early
> >>> branches in the tree of life of modern birds". Science. 346 (6215): 1320–
> >>> 1331. doi:10.1126/science.1253451. PMC 4405904free to read. PMID 25504713.
> >
> > This url takes you right there:
> > http://science.sciencemag.org/content/346/6215/1320.full
> >
> >>> As is so often the case, the latest analysis is made to sound like it
> >>> settles the question of phylogeny, but I have to wonder. "et al"
> >>> includes something like seventy (!) co-authors, which is not always
> >>> conducive to a consistent article.
> >
> > The first sentence in the abstract is already imprecise:
> >
> > To better determine the history of modern birds, we performed a
> > genome-scale phylogenetic analysis of 48 species representing
> > all orders of Neoaves using phylogenomic methods created to handle
> > genome-scale data.
>
> > It also includes some representatives of some (not all) orders of
> > Palaeognathae and Galloanserae, presumably to root the tree.
> > You can find them in Fig. 1:
> > https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg
>
> Is this something more than a nitpick? Do you draw some lesson?

Just that it would have been good to put something like my
sentence immediately afterwards. For one thing, Fig. 1 gives
estimates of divergence times for the orders involved, two from
each of Palaeognathae and Galloanserae, and also
all other divergence times that this inclusion entails.

The deepest, Palaeognathae diverging from Neognathae, is put at
more than 100mya. That predates those iconic Cretaceous birds,
Hesperornis and Ichthyornis.

> >> Turns out much of it can be supported by independent analyses.
> >>
> >>> That aside, the abstract also
> >>> sounds a sort of cautionary note:
> >>>
> >>> Even with whole genomes, some of the earliest branches in Neoaves proved
> >>> challenging to resolve, which was best explained by massive protein-coding
> >>> sequence convergence and high levels of incomplete lineage sorting that
> >>> occurred during a rapid radiation after the Cretaceous-Paleogene mass
> >>> extinction event about 66 million years ago.
> >>>
> >>> Then too, I could see no sign of fossil evidence being used in this
> >>> study, although a couple of references do seem to center on the fossil
> >>> record.
> >>
> >> I see no problem with not using fossil evidence in a genomic study.
> >
> > Given the "massive protein-coding sequence convergence," it seems
> > prudent to make sure the fossil evidence is not at odds with the
> > molecular.
>
> So it seems to you. But in fact the best check would be different
> molecular data, which in fact Jarvis et al. did.

Looks to me like the default assumption has become that if morphological
analyses and molecular analyses conflict, the molecular should
prevail. Can you think of a single exception anywhere in the
last decade, matching the best each side had to offer?

> Would you really expect
> every study to use all different sorts of data, gained in completely
> different ways and demanding different sorts of expertise?

No, why would you think that? I would only want some mention of
what others have found by different methods. With about seventy
co-authors, a few could have been detailed to provide such mention.

A very modest addition, which has nothing to do with the phylogeny,
would have been to use a different shade in each branch of Fig. 1
for the times for which we have fossil data.

Three or four years ago, there was a paper in _Science_ about
the phylogeny of placental orders which included a tree very
much like Fig. 1 but which also used different shades along
each branch to show the times for which we have fossil data.
It was really interesting to see at a glance, for example,
how long the gaps in Tubulidentata [aardvarks} were.

> Would you
> also suggest that every paleo paper should feature a DNA sequence analysis?

You seem to be thinking only in terms of reconstructing evolutionary
trees. There is an awful lot more to paleontology than that, otherwise
it would lose most of its attraction to the general public.

"Jurassic Park" or _Dinotopia_ would not have sold well if it had
been mostly about systematics .

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer --

University of South Carolina at Columbia
http://people.math.sc.edu/nyikos/

[John Harshman]

There are no sides, though some people in the business seem to act as if
there were. Not sure what "best" would mean either. The great things
about the molecular data are that 1) it's easy to test your results by
gathering more data, but there are only so many lumps on a bone and and
2) character sampling and coding is simpler and more objective.

Molecules have been winning lately for a reason, and it's not just
prejudice. The days when guinea pigs weren't rodents are long gone. Of
course the conflicts are fairly rare, but when they happen, the
molecular data tend to be both more plentiful and more replicable.

>> Would you really expect
>> every study to use all different sorts of data, gained in completely
>> different ways and demanding different sorts of expertise?
>
> No, why would you think that? I would only want some mention of
> what others have found by different methods. With about seventy
> co-authors, a few could have been detailed to provide such mention.

I suspect that few if any of the co-authors of that paper were familiar
with the morphological literature. If you want a counterweight, try
Livezey & Zusi 2007.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2517308/

But I warn you, it's deeply flawed. There are suspiciously high
consistencies among characters as well as a summary dismissal of
molecular data in the introduction. Mostly, I'm afraid it's an exercise
in confirmation bias. I don't want to speak ill of the dead (Brad
Livezey died in a car accident in 2011), but there are big
methodological problems with much of his work.

> A very modest addition, which has nothing to do with the phylogeny,
> would have been to use a different shade in each branch of Fig. 1
> for the times for which we have fossil data.

I wouldn't make a figure more complicated for such a slight and
confusing reason.

> Three or four years ago, there was a paper in _Science_ about
> the phylogeny of placental orders which included a tree very
> much like Fig. 1 but which also used different shades along
> each branch to show the times for which we have fossil data.
> It was really interesting to see at a glance, for example,
> how long the gaps in Tubulidentata [aardvarks} were.

I'm suspecting the figure was there specifically for that reason, and
didn't combine multiple goals. The main figure in a paper specifically
about presenting a phylogenetic hypothesis is not the place.

>> Would you
>> also suggest that every paleo paper should feature a DNA sequence analysis?
>
> You seem to be thinking only in terms of reconstructing evolutionary
> trees. There is an awful lot more to paleontology than that, otherwise
> it would lose most of its attraction to the general public.

Not sure of the relevance of the attraction to the general public. Or
why this means that every systematics paper should present some
paleontology.

> "Jurassic Park" or _Dinotopia_ would not have sold well if it had
> been mostly about systematics .

Jurassic Park would have been better had they only consulted a
systematist or two. Frog DNA? Seriously?

[John Harshman]

Genetically, there are indeed no living fossils. Molecular evolution
never stops. And the molecular data show that the divergence among
extant species is very recent.

>>>> But genetically,
>>>> extant species of flamingos are all very similar. I would estimate a
>>>> Pleistocene divergence.
>
> Over on a thread about squirrels that Oxyaena began, I quoted the following
> from Carroll [1988]:
>
> Miocene members of the genus *Sciurus* are so similar to
> living squirrels that Emry and Thorington (1984) considered this
> genus to be a "living fossil," despite its worldwide distribution
> and great species diversity in the modern fauna.
>
> Note, "diversity" not "disparity."

What is the relevance to our current discussion?

>>>> More interesting to me is their divergence from
>>>> their living sister group, grebes.
>>>
>>> According to Wikipedia, the extant grebe genus *Podiceps* goes back
>>> at least to the early Miocene, as far back as any true grebe goes.
>>> But FWIW, the Jarvis article puts the divergence much further back,
>>> to 55 mya.
>>
>> I'm talking about the divergence within crown flamingos.
>
> That was before you talked about "their divergence from
> their living sister group, grebes." You want the crown
> group including both flamingos and grebes for that.

True.

>>> Keep in mind that "molecular clock" estimates are much less
>>> precise than phylogenetic estimates. It was only a few years ago
>>> that the molecular estimates for the LCA of Placentalia were brought
>>> into somewhat closer line with the fossil evidence.
>>>
>>>> Gerald Mayr had a paper about this,
>>>> including a possibly basal flamingo:
>>>>
>>>> http://www.senckenberg.de/files/content/forschung/abteilung/terrzool/ornithologie/flamingo.pdf
>>>
>>> I note that he puts grebes and flamingos in a clade with loons
>>> and penguins and away from every other bird analyzed, using
>>> morphological evidence. That's a far cry from where Jarvis et. al.
>>> put loons and penguins using molecular evidence. Compare Mayr's
>>> "strict consensus tree" in his Figure 1 with Jarvis's Fig. 1:
>>> https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg
>>
>> Yes. Mayr's analysis is wrong about that. Jarvis is correct.
>
> Was Mayr inconsistent, or do you know of some independent analysis
> (preferably morphological) that confirms Jarvis's correctness?

Not sure what you mean by "inconsistent". Inconsistent with what? Why
"preferably morphological"? There are a number of independent molecular
analyses that confirm Jarvis et al. on this point. Hackett et al. 2008,
for example (though really, the confirmation is the other way around,
Hackett et al. being prior). Morphological analyses have found all
manner of contradictory and poorly supported relationships among avian
orders.

[Peter Nyikos]

That assumes that "molecular clocks" are very reliable. But what kind
of DNA do they use? If the DNA is translated into crucial proteins,
like aa-tRNA synthetases, one can expect it to be very tightly conserved.
See also below what I wrote earlier about Placentalia.

> >>>> But genetically,
> >>>> extant species of flamingos are all very similar. I would estimate a
> >>>> Pleistocene divergence.
> >
> > Over on a thread about squirrels that Oxyaena began, I quoted the following
> > from Carroll [1988]:
> >
> > Miocene members of the genus *Sciurus* are so similar to
> > living squirrels that Emry and Thorington (1984) considered this
> > genus to be a "living fossil," despite its worldwide distribution
> > and great species diversity in the modern fauna.
> >
> > Note, "diversity" not "disparity."
>
> What is the relevance to our current discussion?

Morphological stasis over tens of million years.

> >>>> More interesting to me is their divergence from
> >>>> their living sister group, grebes.
> >>>
> >>> According to Wikipedia, the extant grebe genus *Podiceps* goes back
> >>> at least to the early Miocene, as far back as any true grebe goes.
> >>> But FWIW, the Jarvis article puts the divergence much further back,
> >>> to 55 mya.
> >>
> >> I'm talking about the divergence within crown flamingos.
> >
> > That was before you talked about "their divergence from
> > their living sister group, grebes." You want the crown
> > group including both flamingos and grebes for that.
>
> True.
>
> >>> Keep in mind that "molecular clock" estimates are much less
> >>> precise than phylogenetic estimates. It was only a few years ago
> >>> that the molecular estimates for the LCA of Placentalia were brought
> >>> into somewhat closer line with the fossil evidence.

And some divergence times still seem excessive to me, like the
metatheria-eutheria split.

> >>>> Gerald Mayr had a paper about this,
> >>>> including a possibly basal flamingo:
> >>>>
> >>>> http://www.senckenberg.de/files/content/forschung/abteilung/terrzool/ornithologie/flamingo.pdf
> >>>
> >>> I note that he puts grebes and flamingos in a clade with loons
> >>> and penguins and away from every other bird analyzed, using
> >>> morphological evidence. That's a far cry from where Jarvis et. al.
> >>> put loons and penguins using molecular evidence. Compare Mayr's
> >>> "strict consensus tree" in his Figure 1 with Jarvis's Fig. 1:
> >>> https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg
> >>
> >> Yes. Mayr's analysis is wrong about that. Jarvis is correct.
> >
> > Was Mayr inconsistent, or do you know of some independent analysis
> > (preferably morphological) that confirms Jarvis's correctness?
>
> Not sure what you mean by "inconsistent". Inconsistent with what?

Was one part of the analysis inconsistent with some other part?

> Why "preferably morphological"?

Because Mayr's was morphological. As long as the morphological
disagrees with the molecular, there will be a problem unless
someone announces that the molecular estimates are "settled
science, and so debate must end" and has enough influence to
make his announcement "the consensus" [whatever that means].

> There are a number of independent molecular
> analyses that confirm Jarvis et al. on this point. Hackett et al. 2008,
> for example (though really, the confirmation is the other way around,
> Hackett et al. being prior). Morphological analyses have found all
> manner of contradictory and poorly supported relationships among avian
> orders.

Now it's my turn to ask you, "Contradictory with what?"

Peter Nyikos
Professor, Dept. of Mathematics -- standard disclaimer --

The original USC
http://people.math.sc.edu/nyikos/

[John Harshman]

No it doesn't. It merely assumes they are not off by a factor of 10 or more.

> But what kind
> of DNA do they use? If the DNA is translated into crucial proteins,
> like aa-tRNA synthetases, one can expect it to be very tightly conserved.
> See also below what I wrote earlier about Placentalia.

The initial result comes from Sibley & Ahlquist, doing DNA-DNA
hybridization of the entire single-copy fraction of the genome, which
would be nearly 100% junk.

>>>>>> But genetically,
>>>>>> extant species of flamingos are all very similar. I would estimate a
>>>>>> Pleistocene divergence.
>>>
>>> Over on a thread about squirrels that Oxyaena began, I quoted the following
>>> from Carroll [1988]:
>>>
>>> Miocene members of the genus *Sciurus* are so similar to
>>> living squirrels that Emry and Thorington (1984) considered this
>>> genus to be a "living fossil," despite its worldwide distribution
>>> and great species diversity in the modern fauna.
>>>
>>> Note, "diversity" not "disparity."
>>
>> What is the relevance to our current discussion?
>
> Morphological stasis over tens of million years.

Again, what is the relevance to our current discussion?

>>>>>> More interesting to me is their divergence from
>>>>>> their living sister group, grebes.
>>>>>
>>>>> According to Wikipedia, the extant grebe genus *Podiceps* goes back
>>>>> at least to the early Miocene, as far back as any true grebe goes.
>>>>> But FWIW, the Jarvis article puts the divergence much further back,
>>>>> to 55 mya.
>>>>
>>>> I'm talking about the divergence within crown flamingos.
>>>
>>> That was before you talked about "their divergence from
>>> their living sister group, grebes." You want the crown
>>> group including both flamingos and grebes for that.
>>
>> True.
>>
>>>>> Keep in mind that "molecular clock" estimates are much less
>>>>> precise than phylogenetic estimates. It was only a few years ago
>>>>> that the molecular estimates for the LCA of Placentalia were brought
>>>>> into somewhat closer line with the fossil evidence.
>
> And some divergence times still seem excessive to me, like the
> metatheria-eutheria split.

Once more, what is the relevance to our current discussion?

>>>>>> Gerald Mayr had a paper about this,
>>>>>> including a possibly basal flamingo:
>>>>>>
>>>>>> http://www.senckenberg.de/files/content/forschung/abteilung/terrzool/ornithologie/flamingo.pdf
>>>>>
>>>>> I note that he puts grebes and flamingos in a clade with loons
>>>>> and penguins and away from every other bird analyzed, using
>>>>> morphological evidence. That's a far cry from where Jarvis et. al.
>>>>> put loons and penguins using molecular evidence. Compare Mayr's
>>>>> "strict consensus tree" in his Figure 1 with Jarvis's Fig. 1:
>>>>> https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg
>>>>
>>>> Yes. Mayr's analysis is wrong about that. Jarvis is correct.
>>>
>>> Was Mayr inconsistent, or do you know of some independent analysis
>>> (preferably morphological) that confirms Jarvis's correctness?
>>
>> Not sure what you mean by "inconsistent". Inconsistent with what?
>
> Was one part of the analysis inconsistent with some other part?

What would "parts of the analysis" constitute? Are you asking me whether
there was any hompoplasy on his favored tree? Certainly. A data set with
zero homoplasy would be very suspicious (except for a very few rare
sorts of data).

>> Why "preferably morphological"?
>
> Because Mayr's was morphological. As long as the morphological
> disagrees with the molecular, there will be a problem unless
> someone announces that the molecular estimates are "settled
> science, and so debate must end" and has enough influence to
> make his announcement "the consensus" [whatever that means].

I don't see that as being sensible. You make an artificial division of
data into two camps. If multiple independent molecular data sets agree,
why would you suspend judgment until there was a morphological data set
that agreed too? You also seem unduly interested in authority. That
isn't how modern systematics works. Now there may be a problem in the
sense that some people refuse to believe the results of molecular
analyses, but that's only a problem for those people, not for science.
One may hope you aren't one of those people.

>> There are a number of independent molecular
>> analyses that confirm Jarvis et al. on this point. Hackett et al. 2008,
>> for example (though really, the confirmation is the other way around,
>> Hackett et al. being prior). Morphological analyses have found all
>> manner of contradictory and poorly supported relationships among avian
>> orders.
>
> Now it's my turn to ask you, "Contradictory with what?"

Mutually contradictory.

[Popping mad]

On Tue, 23 Aug 2016 18:15:53 -0700, Peter Nyikos wrote:

> It also includes some representatives of some (not all) orders of
> Palaeognathae and Galloanserae, presumably to root the tree.
> You can find them in Fig. 1:
> https://d2ufo47lrtsv5s.cloudfront.net/content/sci/346/6215/1320/F1.large.jpg

I thought almost all birds are related to ducks.

[John Harshman]

All birds are related to ducks. All birds are related to elephants too.
All life is related to all life.

[Peter Nyikos]

A factor of 3 would already move us from the Plesitocene to well within
the Miocene:

The Pleistocene is the geological epoch which lasted from about 2,588,000
to 11,700 years ago,

The Miocene (pronunciation: /ˈmaɪəˌsiːn/;) is the first geological
epoch of the Neogene Period and extends from about 23.03 to 5.332 million
years ago (Ma).

Both are taken from Wikipedia, and I didn't even have to look
at the articles -- the Google search page already included this
information.

> > But what kind
> > of DNA do they use? If the DNA is translated into crucial proteins,
> > like aa-tRNA synthetases, one can expect it to be very tightly conserved.
> > See also below what I wrote earlier about Placentalia.
>
> The initial result comes from Sibley & Ahlquist, doing DNA-DNA
> hybridization of the entire single-copy fraction of the genome, which
> would be nearly 100% junk.

What is meant by the "single-copy fraction"? Does it include "junk DNA"?

> >>>>>> But genetically,
> >>>>>> extant species of flamingos are all very similar. I would estimate a
> >>>>>> Pleistocene divergence.
> >>>
> >>> Over on a thread about squirrels that Oxyaena began, I quoted the following
> >>> from Carroll [1988]:
> >>>
> >>> Miocene members of the genus *Sciurus* are so similar to
> >>> living squirrels that Emry and Thorington (1984) considered this
> >>> genus to be a "living fossil," despite its worldwide distribution
> >>> and great species diversity in the modern fauna.
> >>>
> >>> Note, "diversity" not "disparity."
> >>
> >> What is the relevance to our current discussion?
> >
> > Morphological stasis over tens of million years.
>
> Again, what is the relevance to our current discussion?

I forgot: you are so completely sold on the idea that
there are no molecular "living fossils" that the idea of
morphological stasis is a non-starter for you.

But perhaps you might consider the following possibility:
despite all the molecular changes which you postulate,
some Miocene squirrels may have had genomes which did not preclude
mating with a squirrel with a Holocene genome and having
fertile offspring.

And hence, they would have been of the same species, according
to one of the most demanding definitions of "same species."

> >>>>>> More interesting to me is their divergence from
> >>>>>> their living sister group, grebes.
> >>>>>
> >>>>> According to Wikipedia, the extant grebe genus *Podiceps* goes back
> >>>>> at least to the early Miocene, as far back as any true grebe goes.
> >>>>> But FWIW, the Jarvis article puts the divergence much further back,
> >>>>> to 55 mya.
> >>>>
> >>>> I'm talking about the divergence within crown flamingos.
> >>>
> >>> That was before you talked about "their divergence from
> >>> their living sister group, grebes." You want the crown
> >>> group including both flamingos and grebes for that.
> >>
> >> True.
> >>
> >>>>> Keep in mind that "molecular clock" estimates are much less
> >>>>> precise than phylogenetic estimates. It was only a few years ago
> >>>>> that the molecular estimates for the LCA of Placentalia were brought
> >>>>> into somewhat closer line with the fossil evidence.
> >
> > And some divergence times still seem excessive to me, like the
> > metatheria-eutheria split.
>
> Once more, what is the relevance to our current discussion?

You lost me: what do you *think* "our current discussion" is about,
besides the one for which "Again, why Pleistocene?" was the
jumping off point. [Well, yeah, there is the divergence between
flamingos and grebes, but look at what a non-starter you made of that.]

Remainder deleted, to be replied to later if the foregoing points
ever get clarified.

Peter Nyikos
Professor of Mathematics -- standard disclaimer --
University of South Carolina
http://people.math.sc.edu/nyikos/

[John Harshman]

Sure, for definitions of "well within" equivalent to a couple of million
years at most. I don't actually think that a factor of 3 is credible in
the current case.

>>> But what kind
>>> of DNA do they use? If the DNA is translated into crucial proteins,
>>> like aa-tRNA synthetases, one can expect it to be very tightly conserved.
>>> See also below what I wrote earlier about Placentalia.
>>
>> The initial result comes from Sibley & Ahlquist, doing DNA-DNA
>> hybridization of the entire single-copy fraction of the genome, which
>> would be nearly 100% junk.
>
> What is meant by the "single-copy fraction"? Does it include "junk DNA"?

It includes everything that's not repetitive, the exact meaning of
"repetitive" depending on exactly how the experiment was conducted. I
suspect that anything present in just two copies was probably fairly
well represented in the sample. So yeah, most of it would be junk DNA.
It's likely that Alu sequences, for example, were almost completely removed.

>>>>>>>> But genetically,
>>>>>>>> extant species of flamingos are all very similar. I would estimate a
>>>>>>>> Pleistocene divergence.
>>>>>
>>>>> Over on a thread about squirrels that Oxyaena began, I quoted the following
>>>>> from Carroll [1988]:
>>>>>
>>>>> Miocene members of the genus *Sciurus* are so similar to
>>>>> living squirrels that Emry and Thorington (1984) considered this
>>>>> genus to be a "living fossil," despite its worldwide distribution
>>>>> and great species diversity in the modern fauna.
>>>>>
>>>>> Note, "diversity" not "disparity."
>>>>
>>>> What is the relevance to our current discussion?
>>>
>>> Morphological stasis over tens of million years.
>>
>> Again, what is the relevance to our current discussion?
>
> I forgot: you are so completely sold on the idea that
> there are no molecular "living fossils" that the idea of
> morphological stasis is a non-starter for you.

You forgot because it isn't true. Morphological stasis is certainly
possible, though I think its prevalence is commonly overstated. That has
almost nothing to do with molecular "living fossils".

> But perhaps you might consider the following possibility:
> despite all the molecular changes which you postulate,
> some Miocene squirrels may have had genomes which did not preclude
> mating with a squirrel with a Holocene genome and having
> fertile offspring.

Quite possible. After all, I know of animals with Miocene common
ancestors (or earlier, perhaps even Eocene) that can do that trick. Of
course there's no way to know except by comparing the divergence dates
of extant species.

> And hence, they would have been of the same species, according
> to one of the most demanding definitions of "same species."

Not a definition that can be applied very well to extant populations,
though. As I may have mentioned before, under that definition there is
one species of Anatidae.

>>>>>>>> More interesting to me is their divergence from
>>>>>>>> their living sister group, grebes.
>>>>>>>
>>>>>>> According to Wikipedia, the extant grebe genus *Podiceps* goes back
>>>>>>> at least to the early Miocene, as far back as any true grebe goes.
>>>>>>> But FWIW, the Jarvis article puts the divergence much further back,
>>>>>>> to 55 mya.
>>>>>>
>>>>>> I'm talking about the divergence within crown flamingos.
>>>>>
>>>>> That was before you talked about "their divergence from
>>>>> their living sister group, grebes." You want the crown
>>>>> group including both flamingos and grebes for that.
>>>>
>>>> True.
>>>>
>>>>>>> Keep in mind that "molecular clock" estimates are much less
>>>>>>> precise than phylogenetic estimates. It was only a few years ago
>>>>>>> that the molecular estimates for the LCA of Placentalia were brought
>>>>>>> into somewhat closer line with the fossil evidence.
>>>
>>> And some divergence times still seem excessive to me, like the
>>> metatheria-eutheria split.
>>
>> Once more, what is the relevance to our current discussion?
>
> You lost me: what do you *think* "our current discussion" is about,
> besides the one for which "Again, why Pleistocene?" was the
> jumping off point. [Well, yeah, there is the divergence between
> flamingos and grebes, but look at what a non-starter you made of that.]

How does a particular error bar for Placentalia, a quite old divergence,
have any relevance for the error bar of crown flamingos? Are you saying
that if one molecular date is off (or, really, "seems excessive to me"),
that's evidence that all of them are?

[Popping mad]

On Thu, 01 Sep 2016 06:57:58 -0700, John Harshman wrote:

>> I thought almost all birds are related to ducks.
>>
> All birds are related to ducks. All birds are related to elephants too.
> All life is related to all life.

right but I thought ducks were the stem group that survived the KP impact

[John Harshman]

That is not correct.